What a giant load of horseshit. You really have absolutely zero idea how all of this works, don't you?

And people making predictions about those booms and busts have been wrong pretty much forever.

DeepSeek is doing exactly that.

DeepSeek is 30 times cheaper than OpenAI and delivers better results.

The reasons DeepSeek is cheaper are its use of less energy and the smart use of prompt caching.

Cuba gets slammed by a lot of hurricanes, which are bad for both solar panels and wind turbines.

I think this is a test but I think it's all about advertising.

Wouldn't it be great to look up into the night sky and see "Coke is It!" glimmering up there. No, no it wouldn't.

How much in additional taxes or airport fees are you willing to pay for that?

Pray tell, what the hell did you think they were going to do with all that data? Do you pay for instagram? No? Then you aren't their customer. You're inventory.. Someone has to pay for all those millions of servers and those terawatts of electricity. If you aren't paying, then you're the one being whored out. Quit clutching your pearls. You folks volunteered for this..

DeepSeek is designing a tensor processor. The instruction set is very simple. Mostly just lots and lots of parallelized low-precision multiply-and-accumulate.

LLMs and humans are both unreliable, but make different mistakes.

Use one to check the other.

Disclaimer: I have used Mechanical Turk many times, but not recently.

Except the summary says that earlier acquisition is more effective than learning "recently".

Most children don't learn a second language by choice, but because they are in a multilingual environment. My daughter is bilingual because from birth to age 10 she shared a bedroom with her monolingual Mandarin-speaking grandmother.

I had an RPN calculator in high school, so that made learning Japanese much easier.

When I was in the military, I was deployed to the Western Pacific. I learned Japanese and Tagalog so I could talk to girls.

Of course they will.

Clients need to do their research, choose the tool they want, and then choose the consultant to help them implement it.

That's the way it works. Microsoft isn't the right partner to help you implement a solution based on Google Gemini. Duh.

Or this could be an example of Jevons Paradox .

As AI makes employees more productive, and thus more profitable, would you fire them or hire more?

Everyone was an essential worker during COVID. Work is how people support their families. Fuck you, elitist.